Predictive Price Model for Buy-Sell Bitcoin

Abstract

This study aims to find a suitable forecasting model for buying and selling bitcoin by using external factors of finance and commodities which can be divided into five major groups: foreign exchange rates, safe assets, crude oil price, the volume of buying and selling bitcoin, and stock market price index. This study used bitcoin price data collected from January 1, 2016 to March 31, 2018. However, we will consider only the latter data collected after the highest value of bitcoin, from December 27, 2017 to October 31, 2018 and we use analysis in the form of multiple linear regression, factor analysis and the Box-Cox transformation theory. When analyzing the data, it was found that    is a better forecasting equation than other forecasting equations when forecasting for a short period of time only. The first factor is Chinese and South Korean stock market price index, safe assets and the foreign exchange rate. The second factor is oil prices and the volume of buying and selling of bitcoin. The third factor is Japan and Germany stock market price index and the fourth factor is the stock market price index for the U.S., Canada and Australia.